Correlating page loading times and conversion rates?
 

Hi

Have a couple of quick questions:


1. Do you keep a tab on your website's performance and correlate it with conversion rates? I mean, to see if any poor performance is affecting your website's session length, bounce rates, conversions, etc? If yes, how do you do that?


2. If and when your website is slow, how do you decide what is the right loading time target for the slow loading page(s)?


Thx

Re: Correlating page loading times and conversion rates?
 

Here's a link that discusses load times in relation to conversions.
https://blog.kissmetrics.com/loading-time/

As to tracking tools, use Google Analytics and set it up to track conversions.

Here are some sites you can use to test your sites performance.

https://gtmetrix.com/
https://tools.pingdom.com/
https://www.webpagetest.org/
https://developers.google.com/speed/pagespeed/insights/

Re: Correlating page loading times and conversion rates?
 

Hi

I recommend that you install and use Google Analytics. It has great tools built in that will allow you to compare and correlate page load times with conversions and other behavior metrics.

Page Timings Report
User Timings Report
Speed Suggestions Report

HTH,

Don Burk

RE: Correlating page loading times and conversion rates?
 

Hi Don
Thanks for the reply - I did attempt to get answers to my Qs via my GA setup prior to posting here. Specific issues I had with looking at GA graphs available via Behavior > Site Speed section of GA:

- Page Timings are available at the granularity of Day, Week, Month (not even hourly). Unfortunately, a day's average in loading time does not expose the sort of trend or fluctuations that I could directly relate with conversions (unless website is performing badly for entire day-week)

- Avg Page Load Time (sec) for many days is 0 sec. From my understanding GA docs, no additional setup is needed to get page loading stats. While my stats on other sections such as sessions, page views, bounce rate seem accurate - my page load time is definitely incorrect.

Thanks for your time!

Edit----------------
After posting the above, I spent more time around GA to understand what could be causing incorrect GA loading time figures. It appears that default site speed sample rate is 1%. I have now set it to 100% - will let it run for a couple of days and see if it changes the performance stats appearing on GA.

Thanks!

RE: Correlating page loading times and conversion rates?
 

Hi CDNgine,
Wrt GA - I have not been able to extract page loading times from GA - please see my response to Don's comment below.

I'll look at gtmetrix but pingdom and webpagetest primarily offer synthetic tests to improve performance. What I'm after is to find something like this for my own site by monitoring actual users:

Hour Load time Bounce Rate
10 2 sec 35%
11 2.5 sec 40%
12 5 sec 55%
13 3 sec 40%

Above figures are of course, made up numbers. But I'm keen to know if actual loading times can be correlated to conversion stats.

Re: Correlating page loading times and conversion rates?
 

1. Yes I keep track but as for correlating it, I honestly don't do it most of the times. This is a given for me at this point - that faster sites will keep customers longer who will transact more. Additionally, as an engineer, having to keep pages fast and snappy is a goal in itself already.

2. Two rules for me - look for low hanging fruit AND keep the common cases fast.

RE: Correlating page loading times and conversion rates?
 

Ofcourse, a general rule that faster the site, better the conversion always prevails.

But, I'm talking about uncovering devil in the details:

So, if I see can track something like - users with page loading time of 3 additional seconds see 10% drop in session length - when I analyze those users, I realize (for example) most of these are Android-Firefox users with certain viewport size.

I can then focus on improving timings for this specific use case knowing the possible gain from this improvement.